Eric T Bradlow

Eric T Bradlow
  • The K.P. Chao Professor
  • Professor of Marketing
  • Faculty Director - Wharton Customer Analytics Initiative
  • Chairperson, Wharton Marketing Department
  • Professor of Economics; Professor of Education; Professor of Statistics

Contact Information

  • office Address:

    761 Jon M. Huntsman Hall
    3730 Walnut Street
    University of Pennsylvania
    Philadelphia, PA 19104

Research Interests: Bayesian computation, latent variable models, marketing research methods, missing data problems, analytics, psychometrics

Links: CV, Twitter

Overview

Professor Eric T. Bradlow is the K.P. Chao Professor, Professor of Marketing, Statistics, Education and Economics and Faculty Director of the Wharton Customer Analytics Initiative. An applied statistician, Professor Bradlow uses high-powered statistical models to solve problems on everything from Internet search engines to product assortment issues. Specifically, his research interests include Bayesian modeling, statistical computing, and developing new methodology for unique data structures with application to business problems.

Eric was recently named a fellow of the American Statistical Association, American Educational Research Association, is past chair of the American Statistical Association Section on Statistics in Marketing, past Editor-in-Chief of Marketing Science, is a past statistical fellow of Bell Labs, and worked at DuPont Corporation’s Corporate Marketing and Business Research Division and the Educational Testing Service.

A prolific scholar, Professor Bradlow’s research has been published in top-tier academic journals such as the Journal of the American Statistical Association, Psychometrika, Statistica Sinica, Chance, Marketing Science, Management Science, and Journal of Marketing Research. He also serves as Associate Editor for the Journal of the American Statistical Association and the Journal of Marketing Research, and is on the Editorial Boards of Marketing Letters, Marketing Science, Journal of Marketing Research, Quantitative Marketing and Economics, and the Quarterly Journal of Electronic Commerce.

Professor Bradlow has won numerous teaching awards at Wharton, including the Anvil Award, MBA Core Curriculum teaching award, the Miller-Sherrerd MBA Core Teaching award and the Excellence in Teaching Award. His teaching interests include courses in Statistics, Marketing Research, Marketing Management and PhD Data Analysis, as well as any material related to customer analytics.

Professor Bradlow earned his PhD and Master’s degrees in Mathematical Statistics from Harvard University and his BS in Economics from the University of Pennsylvania.

Continue Reading

Research

  • Daniel Zantedeschi, Elea McDonnell Feit, Eric Bradlow (2017), Measuring Multi-Channel Advertising Response, Management Science, 63 (8), pp. 2706-2708.

    Abstract: Advances in data collection have made it increasingly easy to collect information on advertising exposures. However, translating this seemingly rich data into measures of advertising response has proven difficult, largely because of concerns that advertisers target customers with a higher propensity to buy or increase advertising during periods of peak demand. We show how this problem can be addressed by studying a setting where a firm randomly held out customers from each campaign, creating a sequence of randomized field experiments that mitigates (many) potential endogeneity problems. Exploratory analysis of individual holdout experiments shows positive effects for both email and catalog; however, the estimated effect for any individual campaign is imprecise, because of the small size of the holdout. To pool data across campaigns, we develop a hierarchical Bayesian model for advertising response that allows us to account for individual differences in purchase propensity and marketing response. Building on the traditional ad-stock framework, we are able to estimate separate decay rates for each advertising medium, allowing us to predict channel-specific short- and long-term effects of advertising and use these predictions to inform marketing strategy. We find that catalogs have substantially longer-lasting impact on customer purchase than emails. We show how the model can be used to score and target individual customers based on their advertising responsiveness, and we find that targeting the most responsive customers increases the predicted returns on advertising by approximately 70% versus traditional recency, frequency, and monetary value–based targeting.

  • Tong Lu, Eric Bradlow, J. Wesley Hutchinson, Binge Consumption of Online Content.

  • Julie Novak, Eleanor McDonnell Feit, Shane T. Jensen, Eric Bradlow (Working), Bayesian Imputation for Anonymous Visits.

  • Valeria Stourm, Eric Bradlow, Peter Fader (2015), Stockpiling Points in Linear Loyalty Programs, Journal of Marketing Research, 52 (2), pp. 253-267.

    Abstract: Customers often stockpile reward points in linear loyalty programs (i.e., programs that do not explicitly reward stockpiling) despite several economic incentives against it (e.g., the time value of money). The authors develop a mathematical model of redemption choice that unites three explanations for why customers seem to be motivated to stockpile on their own, even though the retailer does not reward them for doing so. These motivations are economic (the value of forgone points), cognitive (nonmonetary transaction costs), and psychological (customers value points differently than cash). The authors capture the psychological motivation by allowing customers to book cash and point transactions in separate mental accounts. They estimate the model on data from an international retailer using Markov chain Monte Carlo methods and accurately forecast redemptions during an 11-month out-of-sample period. The results indicate substantial heterogeneity in how customers are motivated to redeem and suggest that the behavior in the data is driven mostly by cognitive and psychological incentives.      

  • Yao Zhang, Eric Bradlow, Dylan Small (2015), Predicting Customer Value Using Clumpiness: From RFM to RFMC, Marketing Science, 34 (2), pp. 195-208.

  • P. Wang, Eric Bradlow, Edward I. George (2014), Meta-Analyses Using Information Reweighting: An Application to Online Advertising, Quantitative Marketing and Economics, 12 (2), pp. 209-233.

  • Eric Schwartz, Eric Bradlow, Peter Fader (2014), Model Selection Using Database Characteristics: Developing a Classification Tree for Longitudinal Incidence Data, Marketing Science , 33 (2), pp. 188-205.

    Abstract: When managers and researchers encounter a data set, they typically ask two key questions: (1) Which model (from a candidate set) should I use? And (2) if I use a particular model, when is it going to likely work well for my business goal? This research addresses those two questions and provides a rule, i.e., a decision tree, for data analysts to portend the “winning model” before having to fit any of them for longitudinal incidence data. We characterize data sets based on managerially relevant (and easy-to-compute) summary statistics, and we use classification techniques from machine learning to provide a decision tree that recommends when to use which model. By doing the “legwork” of obtaining this decision tree for model selection, we provide a time-saving tool to analysts. We illustrate this method for a common marketing problem (i.e., forecasting repeat purchasing incidence for a cohort of new customers) and demonstrate the method’s ability to discriminate among an integrated family of a hidden Markov model (HMM) and its constrained variants. We observe a strong ability for data set characteristics to guide the choice of the most appropriate model, and we observe that some model features (e.g., the “back-and-forth” migration between latent states) are more important to accommodate than are others (e.g., the inclusion of an “off” state with no activity). We also demonstrate the method’s broad potential by providing a general “recipe” for researchers to replicate this kind of model classification task in other managerial contexts (outside of repeat purchasing incidence data and the HMM framework).

  • Arun Gopalakrishnan, Eric Bradlow, Peter Fader (Under Revision), A Cross-Cohort Changepoint Model for Customer-Base Analysis.

    Abstract: We introduce a new methodology that can capture and explain differences across a series of cohorts of new customers in a repeat-transaction setting. More specifically, this new framework, which we call a vector changepoint model, exploits the underlying regime structure in a sequence of acquired customer cohorts, to make predictive statements about new cohorts for which the firm has little or no longitudinal transaction data. To accomplish this, we develop our model within a Hierarchical Bayesian framework to uncover evidence of regime changes for each cohort-level parameter separately, thus disentangling potential explanations for cross-cohort shifts in aggregate transaction patterns.  Calibrating the model using multi-cohort donation data from a non-profit organization, we find that holdout predictions for new cohorts using this model have greater accuracy – and greater diagnostic value – compared to a variety of strong benchmarks. Our modeling approach also highlights the perils of pooling data across cohorts without accounting for cross-cohort shifts, thus enabling managers to quantify their uncertainty about potential regime changes and avoid “old data” aggregation bias.

  • Eric Schwartz, Eric Bradlow, Peter Fader (Under Revision), Customer Acquisition via Display Advertising Using Multi-Armed Bandit Experiments.

    Abstract: Online advertisers regularly deliver several versions of display ads in a single campaign across many websites in order to acquire customers, but they are uncertain about which ads are most effective. As the campaign progresses, they adapt to intermediate results and allocate more impressions to the better performing ads on each website. But how should they decide what percentage of impressions to allocate to each ad?  This paper answers that question, resolving the classic "explore/exploit" tradeoff using multi-armed bandit (MAB) methods. However, this marketing problem contains challenges, such as hierarchical structure (ads within a website), attributes of actions (creative elements of an ad), and batched decisions (millions of impressions at a time), that are not fully accommodated by existing MAB methods. We address this marketing problem by utilizing a hierarchical generalized linear model with unobserved heterogeneity combined with an algorithm known as Thompson Sampling. Our approach captures how the impact of observable ad attributes on ad effectiveness differs by website in unobserved ways, and our policy generates allocations of impressions that can be used in practice. We implemented this policy in a live field experiment delivering over 700 million ad impressions in an online display campaign with a large retail bank. Over the course of two months, our policy achieved an 8% improvement in the customer acquisition rate, relative to a control policy, without any additional costs to the bank. Beyond the actual experiment, we performed counter-factual simulations to evaluate a range of alternative model specifications and allocation rules in MAB policies.

  • Eric Bradlow Bradlow Clumpiness Spreadsheet.

Teaching

Current Courses

  • MKTG612 - Dynamic Marketing Strategy

    Building upon Marketing 611, the goal of this course is to develop skills in formulating and implementing marketing strategies for brands and businesses. The course will focus on issues such as the selection of which businesses and segments to compete in, how to allocate resources across businesses, segments, and elements of the marketing mix, as well as other significant strategic issues facing today's managers in a dynamic competitive environment. ,A central theme of the course is that the answer to these strategic problems varies over time depending on the stage of the product life cycle at which marketing decisions are being made. As such, the PLC serves as the central organizing vehicle of the course. We will explore such issues as how to design optimal strategies for the launch of new products and services that arise during the introductory phase, how to maximize the acceleration of revenue during the growth phase, how to sustain and extend profitability during the mature phase, and how to manage a business during the inevitable decline phase.

    MKTG612002 ( Syllabus )

    MKTG612004 ( Syllabus )

    MKTG612006 ( Syllabus )

    MKTG612008 ( Syllabus )

Past Courses

  • MKTG212 - Data and Analysis for Marketing Decisions

    Firms have access to detailed data of customers and past marketing actions. Such data may include in-store and online customer transactions, customer surveys as well as prices and advertising. Using real-world applications from various industries, the goal of the course is to familiarize students with several types of managerial problems as well as data sources and techniques, commonly employed in making effective marketing decisions. The course would involve formulating critical managerial problems, developing relevant hypotheses, analyzing data and, most importantly, drawing inferences and telling convincing narratives, with a view of yielding actionable results.

  • MKTG399 - Independent Study

  • MKTG611 - Marketing Management

    This course addresses how to design and implement the best combination of marketing efforts to carry out a firm's strategy in its target markets. Specifically, this course seeks to develop the student's (1) understanding of how the firm can benefit by creating and delivering value to its customers, and stakeholders, and (2) skills in applying the analytical concepts and tools of marketing to such decisions as segmentation and targeting, branding, pricing, distribution, and promotion. The course uses lectures and case discussions, case write-ups, student presentations, and a comprehensive final examination to achieve these objectives.

  • MKTG612 - Dynamic Marketing Strategy

    Building upon Marketing 611, the goal of this course is to develop skills in formulating and implementing marketing strategies for brands and businesses. The course will focus on issues such as the selection of which businesses and segments to compete in, how to allocate resources across businesses, segments, and elements of the marketing mix, as well as other significant strategic issues facing today's managers in a dynamic competitive environment. ,A central theme of the course is that the answer to these strategic problems varies over time depending on the stage of the product life cycle at which marketing decisions are being made. As such, the PLC serves as the central organizing vehicle of the course. We will explore such issues as how to design optimal strategies for the launch of new products and services that arise during the introductory phase, how to maximize the acceleration of revenue during the growth phase, how to sustain and extend profitability during the mature phase, and how to manage a business during the inevitable decline phase.

  • MKTG613 - Strategic Marketing Simulation

    Building upon Marketing 611, Marketing 613 is an intensive immersion course designed to develop skills in formulating and implementing marketing strategies for brands and businesses. The central activity will be participation in a realistic integrative product management simulation named SABRE. In SABRE, students will form management teams that oversee all critical aspects of modern product management: the design and marketing of new products, advertising budgeting and design, sales force sizing and allocation, and production planning. As in the real world, teams will compete for profitability, and the success that each team has in achieving this goal will be a major driver of the class assessment. ,The SABRE simulation is used to convey the two foci of learning in the course: the changing nature of strategic problems and their optimal solutions as industries progress through the product life cycle, and exposure to the latest analytic tools for solving these problems. Specifically, SABRE management teams will receive training in both how to make optimal use of marketing research information to reduce uncertainty in product design and positioning, as well as decision support models to guide resource allocation.

  • MKTG899 - Independent Study

    A student contemplating an independent study project must first find a faculty member who agrees to supervise and approve the student's written proposal as an independent study (MKTG 899). If a student wishes the proposed work to be used to meet the ASP requirement, he/she should then submit the approved proposal to the MBA adviser who will determine if it is an appropriate substitute. Such substitutions will only be approved prior to the beginning of the semester.

  • MKTG956 - Empirical Models in Marketing - Part A

    This course is designed to generate awareness and appreciation of the way several substantive topics in marketing have been studied empirically using quantitative models. This seminar reviews empirical models of marketing phenomena including consumer choice, adoption of new products, sales response to marketing mix elements, and competitive interaction. Applies methods and concepts developed in econometrics and statistics but focuses on substantive issues of model structure and interpretation, rather than on estimation techniques. Ultimately, the goals are a) to prepare students to read and understand the literature and b) to stimulate new research interests. By the end of the course, students should be familiar with the key issues and approaches in empirical marketing modeling.

  • MKTG957 - Empirical Models in Marketing - Part B

    This course is designed to generate awareness and appreciation of the way several substantive topics in marketing have been studied empirically using quantitative models. This seminar reviews empirical models of marketing phenomena including consumer choice, adoption of new products, sales response to marketing mix elements, and competitive interaction. Applies methods and concepts developed in econometrics and statistics but focuses on substantive issues of model structure and interpretation, rather than on estimation techniques. Ultimately, the goals are a) to prepare students to read and understand the literature and b) to stimulate new research interests. By the end of the course, students should be familiar with the key issues and approaches in empirical marketing modeling.

  • MKTG995 - Dissertation

  • MKTG999 - Supervised Independent Study

    Requires written permission of instructor and the department graduate adviser.

  • STAT101 - Introductory Business Statistics

    Data summaries and descriptive statistics; introduction to a statistical computer package; Probability: distributions, expectation, variance, covariance, portfolios, central limit theorem; statistical inference of univariate data; Statistical inference for bivariate data: inference for intrinsically linear simple regression models. This course will have a business focus, but is not inappropriate for students in the college.

  • STAT111 - Introductory Statistics

    Introduction to concepts in probability. Basic statistical inference procedures of estimation, confidence intervals and hypothesis testing directed towards applications in science and medicine. The use of the JMP statistical package.

  • STAT500 - Applied Regression and Analysis of Variance

    An applied graduate level course in multiple regression and analysis of variance for students who have completed an undergraduate course in basic statistical methods. Emphasis is on practical methods of data analysis and their interpretation. Covers model building, general linear hypothesis, residual analysis, leverage and influence, one-way anova, two-way anova, factorial anova. Primarily for doctoral students in the managerial, behavioral, social and health sciences.

  • STAT501 - Introduction to Nonparametric Methods and Log-linear Models

    An applied graduate level course for students who have completed an undergraduate course in basic statistical methods. Covers two unrelated topics: loglinear and logit models for discrete data and nonparametric methods for nonnormal data. Emphasis is on practical methods of data analysis and their interpretation. Primarily for doctoral students in the managerial, behavioral, social and health sciences. May be taken before STAT 500 with permission of instructor.

Awards and Honors

  • J.B. Steenkamp Long-Term Impact Award, IJRM Best Paper, 2016
  • Anvil Award, Best Teacher in Wharton MBA Program, 2015
  • Inaugural Fellow of the University of Pennsylvania, 2009
  • Fellow of the American Education Research Association, 2009
  • Finalist, H. Paul Root Award, Best Paper in Journal of Marketing, 2009
  • Finalist, John D.C. Little Best Paper Award, 2008
  • Wharton East WEMBA Teaching Award, 2008
  • American Marketing Association EXPLOR Award, 2007
  • Wharton East WEMBA Teaching Award, 2007
  • Wharton School, MBA Excellence in Teaching Award, 2002, 2003, 2004, 2005, 2006, 2006 Description

    Wharton School, MBA Excellence in Teaching Award

  • NCME Technical or Scientific Contribution to the Field of Educational Measurement: Development of Testlet Response Theory, 2006
  • Wharton East WEMBA Teaching Award, 2006
  • Wharton School, MBA Excellence in Teaching Award, 2003, 2004, 2005, 2006, 2006
  • “Goes Above and Beyond the Call of Duty” Wharton MBA Teaching Award, 2006
  • Fellow of the American Statistical Association, 2005 Description

    Fellow of the American Statistical Association

  • First recipient of The K.P. Chao Professorship, 2005 Description

    Named the first recipient of The K.P. Chao Professorship

  • Helen Kardon Moss Anvil Award Finalist, 2001-2002, 2004-2005, 2005
  • Appointed Fellow of the American Statistical Association, 2005
  • Finalist, Paul E. Green Award for the best paper in Journal of Marketing Research, 2004 Description

    “A Learning-based Model for Imputing Missing Levels in Partial Conjoint Profiles,” co-authored with Y. Hue and T-H Ho, lead article and discussion paper, Vol. XLI (November 2004), 369-38

  • Wharton Undergraduate Excellence in Teaching Award, 2004, 2004
  • Wharton School, MBA Excellence in Teaching Award, 2005 Description

    2003, 2004, 2005

  • AERA Outstanding Reviewer, 2003 Description

    AERA Outstanding Reviewer

  • Wharton West WEMBA Teaching Award, 2003 Description

    Wharton West WEMBA Teaching Award

  • AERA Outstanding Reviewer, 2003
  • Helen Kardon Moss Anvil Award Finalist, 2005 Description

    2001-2002, 2004-2005

  • Miller-Sherrerd MBA Core Teaching Award, 2002 Description

    1999, 2000, 2001, 2002

  • Wharton MBA Core Curriculum Teaching Award, 2001 Description

    1998, 1999, 2001

  • Appointed Research Consultant, AT&T Bell Laboratories, 1997 Description

    Appointed Research Consultant, AT&T Bell Laboratories

  • Finalist, American Statistical Association Savage Award Dissertation Prize, 1997 Description

    Finalist, American Statistical Association Savage Award Dissertation Prize

  • E.I. DuPont de Nemours and Company young researcher award, 1992 Description

    Corporate Marketing Division

  • Harvard University Derek Bok Center for excellence in teaching, 1991 Description

    4-time winner

In the News

Knowledge @ Wharton

Activity

Latest Research

Daniel Zantedeschi, Elea McDonnell Feit, Eric Bradlow (2017), Measuring Multi-Channel Advertising Response, Management Science, 63 (8), pp. 2706-2708.
All Research

In the News

How Will Targeted Ads Fare in an Era of Data Protection?

Although many companies are uncertain how the General Data Protection Regulation that went into effect last month will impact them, one thing is clear: They will not be able to target their advertising as freely as in the past.

Knowledge @ Wharton - 2018/06/22
All News

Awards and Honors

J.B. Steenkamp Long-Term Impact Award, IJRM Best Paper 2016
All Awards